Monday, 22 September 2008


I am no philosopher. I have no great insights. Big questions like, "What is mental illness?" are for greater minds than mine to ponder. But, as a clinician seeing folk day to day, it does fall to me to arbitrate over the implications of what's the business of mental healthcare. Simply put, what do my team and I do?
Other than eating biscuits (liaison nurses have the best ones at the moment), we need to define what work we'll do. We need something more workable than, "make things better, make it good enough, help folk cope," even if that's the reality of it.

So I started musing over what the concept of what mental illness means from a healthcare perspective and what we do and don't see as mental illness.

Comments were very interesting. One specific point was made. Also, two general themes that emerged were :
- what's psychiatric mental illness in contrast to other changes of mental wellbeing?
- what's the relevance of mental disorder and the Mental Health Act 1983 in managing mental health problems, now we've got the Mental Capacity Act 2005?
Those two themes merit a post in their own right, so I'll post about those another day since the comments have started stirring my thoughts in a bit of a whirl and I haven't quite got my head around all the implications yet.

The specific point was on diagnosis. I believe diagnosis, done openly and honestly, can serve a useful purpose (with patients and carers and me finding it useful, as I use it). One suggestion was that treatment (drug and psychological therapy) is guided by symptom clusters rather than a diagnostic label. I'd take that further and think that if you have a symptom cluster that has significance, when the presence of those particular symptoms and the intensity of them are all taken together, then this forms a definition of a clinical syndrome. Which is a diagnostic label. As such, I'm still wedded to the belief that attention to symptoms and the impact of symptoms is the key, but I'd then encapsulate the summary of symptoms (and signs) in a diagnosis.

Does diagnois matter, is it useful, or conversely is it unhelpful? When I was in General Practice diagnosis was managed differently by different GPs. Some were firmly wedded to diagnosis, saw it as a critical role of a GP to diagnose what was wrong, share it with the patient so they had an understanding of what the medical perspective was of their collection of symptoms they presented with, then enabled appropriate treatment for that condition. Other GPs were fuzzy in using diagnosis, they were great clinicians and adeptly sleuthed out what was going on but wouldn't share formal diagnostic terms with patients. Their patients would say, "I see Dr X for my heart condition," or "Dr Y looks after my breathing problems, she keeps my lungs cleared up," and so forth. Mostly this worked well, patients had confidence that their GP was managing their problems, knew what was going on, explained to the patient what was being done but didn't harp on about blah blah blah syndrome which had no meaning to the patient.


Apart from when patients got worse.

"Dr Z is treating my stomach complaint but now I've pain, nausea and am totally off my food, I'm vomiting, I've weight loss, my skin's yellow and I'm really tired all the time." That's because you have stomach cancer with widespread malignancy and liver metastases, you're getting worse, you're going to die, but the diagnosis of "cancer" has never been raised so management of consequences of this have never really been discussed since you can't have that kind of talk for, "a stomach ulcer that's gone a bit nasty," which Dr Z is treating.

Ducking diagnosis, with the issues of problems progressing, yet everyone colluding not to look it in the eye and call it what it is, has generated so much poor clinical care for patients I've seen that I'm really, really wary of treading this path now.

Diagnosis and honesty in what's going on means you can then have meaningful dialogue about what's going on, what the future may hold, what choices are before you and what choices you may want to consider for the future. A disadvantage of discussing symptoms in the here and now (instead of diagnosis) is that it doesn't lend itself to that kind of discussion as readily. It does have the advantage that it can focus on symptom management which can be more positive, presuming that improvement of symptoms and coping so all's good enough is restoration to a state of relative health, so all's good (in contrast to a diagnosis that can linger on forever).

Over the last few weeks I've seen many patients.

One lady had poor concentration and poor attention span, poor organisational skills, poor memory of recent events and low mood with mild anxiety. She had dementia (F00.131 dementia in Alzheimer's disease, late onset, with predominantly depressive symptoms, of moderate severity). She could benefit from an antidementia drug.

One lady had poor concentration and poor attention span, poor organisational skills, poor memory of recent events and low mood with mild anxiety. She had depression (F32.11 moderate depressive episode, with somatic syndrome). She could benefit from CBT and an antidepressant drug.

One lady had poor concentration and poor attention span, poor organisational skills, poor memory of recent events and low mood with mild anxiety. She had a bereavement (F43.22 adjustment disorder, mixed anxiety and depressive reaction). She could benefit from nursing or chaplaincy support.

Symptoms of poor concentration and poor attention span, poor organisational skills, poor memory of recent events and low mood with mild anxiety don't determine the treatment, it's the clinical syndrome (and thus the diagnosis) which frames how we can conceptualise the patients' experiences and generate appropriate interventions to then offer.

Diagnosis can be exceptionally unhelpful, but most of the time I do think it's useful since it can guide more widely on the management of the symptoms and on the likely future course.

Given as a helpful handle and explanation, with specific detail that has meaning, a diagnosis can be important. For example, one type of dementia needs different drugs to other types of dementia, and will run a different course, so rather than "memory problems" or "dementia" being discussed, a diagnosis can be useful in helping us discuss drug treatment, Council Tax exemption, psychological interventions (which are at least as important as drug treatment in maintaining cognition), formally making Advance Decisions, driving and the DVLA, Direct Payments, Lasting Power of Attorney for financial decisions and for health/social welfare decisions and so on.

I'm sold that it's the patients' experience that is at the heart of what we do. As such, symptoms matter. They're what our patient is grappling with, what they're telling us they find hard to manage, they're what impact upon their ability to function/to crack on and do what they would wish to do. The number and the intensity of symptoms (the symptoms burden) is key to understanding. But although this informs our assessment, formulation, interventions and carer support, I truly believe that diagnosis still has some use in adding to this process.


Pem said...

Your point is a good one in the situation you cite, where the same symptoms can come from different diagnoses. What worries me is when those symptom clusters have not been fully differentiated and varying symptom clusters are lumped under one diagnosis. My son has ADD with some autistic features, but I have avoided the Asperger's diagnosis because I think that will lead not to more appropriate treatment but to less appropriate treatment. He doesn't fit the standard pattern of Aspergers.

The other interesting case is very early stages. My husband has been diagnosed with Lewy Body Dementia. He is currently in Washington DC doing research at the Library of Congress and staying with an aunt. I am, in the balance, glad I know and can prepare for the future, and Aricept has helped him. But it is hard knowledge to live with, particularly when almost everyone says "He seems fine."

Paul said...

(Apologies in advance for the long comment)

Regarding diagnosis, you make a very strong case for the absolute importance of this approach in physical health care.

However applying these arguments to mental health without some qualification is problematic in my view. Here are two things which trouble me when thinking about this:

1. The interaction problem

(This one is only going to be an issue for you if you believe, as I do, that psychological factors (beliefs, values, emotions, volition etc) have some role to play in the aetiology and / or maintenance of MH problems.)

The problem in MH (and to some extent in physical health) is that the very condition being diagnosed can improve or worsen in response to being told one's diagnosis. For example, panic attacks can become panic disorder when a person appraises normal psychological and physical phenomena (e.g., thoughts racing, pulse racing) as evidence for impending mental or physical catastrophe (e.g., ‘I’m going mad’ or ‘I’m having a heart attack’). Conversely, being informed you simply have ‘panic disorder’ can reduce your conviction that you are not 'going mad' or having a heart attack. If you have faith in this being the correct diagnosis, then this will reduce your anxiety and reduce the frequency of your panic attacks. There is a beneficial interaction between the diagnois and the condition being diagnosed.

It can go the other way too though. Imagine the following:

Someone comes to you saying that they have started hearing voices and they worry this means they are going mad. Assessment suggests the voices are much more frequent and intense whenever the person gets really anxious. The person only gets really anxious whenever they find themselves worrying that the voices mean they are developing a psychotic illness (such as schizophrenia).

Given the above, we might reasonably predict that diagnosing that person with a psychotic illness is likely to increase their anxiety and in turn increase the frequency and intensity of the voice hearing (notably, risk of suicide also increases following a diagnosis of major mental illness and if fears of mental disintegration are present). This is an example of a detrimental interaction between diagnosis and the condition being diagnosed.

I suppose we could also predict that the less 'psychological' a problem is, the less there is an interaction effect (beneficial or detrimental) Clearly, diagnosing a person with dementia does not cause them to develop dementia.

Whether the interaction effect is an argument against diagnosis in MH, or an argument against telling people their diagnosis is not clear to me. Follow the link to an interesting response by philosopher Thomas Pogge to this very problem:

You’ll note from his response that complete honesty may not always be the most ethical policy.

2. Values

Consideration of values is much more relevant in MH than in physical health. In physical health, it's fairly uncontroversial to refer to problems encountered as 'illnesses' and 'diseases'. We don't disagree over the negative value of cancer, for example.

But in mental health, controversy reigns supreme - whether professionals or service users like it or not. In such a climate, 'objective' diagnoses are a hard call. This is why I think distress (subjectively defined) and impaired capacity are the only decent justifications for psychiatric input. That is, diagnosis on it’s own is not enough. I think this argument holds true across constructs like 'delusions' and 'personality disorder' and so on.

If a client is distressed and asking for help and / or has impaired capacity I agree the benefits of an accurate, comprehensive, reliable and valid diagnosis are immense. Not harming the client with inappropriate treatment, the planning of effective interventions, the prediction and prevention of future distress / impairment and developing better treatments are all very good reasons for retaining some sort of classification system.

However I agree with the view that the ICD-10 / DSM-IV systems DO NOT provide reliable or valid diagnostic categories when it comes to things like depression, anxiety, ‘personality disorder’ or psychosis – primarily because of reification of social norms, astonishingly high levels of comorbidity (i.e., a lack of the desired symptom clusters) and poor reliability.

I may be wrong but for these and some other problems I agree with those who argue the solution is likely to be a dimensional symptom focused approach where the cut-off point between ‘health’ and ‘illness’ is driven by considerations of distress, help-seeking, impaired decision-making capacity and available resources.

Judgements of ‘mental illness’ should no longer be talked about as if they were objective and value free. Instead we should make the value judgements explicit and we should get them right, as far as we can.

The best way to get the values right, in my view, is to promote service user involvement as far as possible.

Adiemus said...

Great post, and relevant to mine today about case formulation!
I don't like 'diagnosis' particularly because I think it leaves us without possible explanations for individual unique presentations - and diagnoses might help you and I talk about what is going on (as a sort of verbal shorthand), but it doesn't unpack the various factors that may need to be addressed during intervention.
I wonder whether a diagnosis is so helpful when we start looking at a biopsychosocial model - or when it's disability we're looking at rather than impairment.
Perhaps diagnoses are great in certain settings (eg research, or where a clear mechanism has been identified, or where clinicians need to communicate with each other) - but less useful during the process of therapy, or where the impact of a problem starts to affect the life of a person.
Bronnie, Adiemusfree